Questions & Answers

Regarding the AN-2000 Flow Measurement Instruments


How do I install the ANFLOW software to my WIN95 system?

In order to properly install ANFLOW for WIN95 upgrade, you have to do the following things:
    1) De-compress the file AN_WIN.EXE after you have loaded it to a separate directory.
    2) The compression program would ask if it can open a "DISK" directory Answer Y.
    3) You will get a directory named "DISK" with a file called "setup.exe".
You must run this file from WIN95.
    4) Upon installation, you will be prompted to enter the COM port which you want to assign to the serial port on the AN-2000 PC card. You must select one of the two options, according to other setups in your PC (modems? Serial printers? Plotters? Network?etc.)
The selection of the COM port will enter the following lines in system.ini file:


COM3Irq=5
COM3Base=01C00
or:


COM4Irq=10
COM4Base=01C00


   5) You should enter the "My computer/Control panel/Add new hardware" screen and add a new port (COM3 or COM4)

   6) Reboot Windows.

   7) Enter "My computer/Control panel/System" and check that the new port was added (in COM \ LPT ports menu). Double click on the COM name that you added (select RESOURCES) and check that the address and IRQ numbers are correct. If yes- go to step 8. If no - change and go to step 6.    8) The installation program will create a directory called ANFLOW.WIN.
You have to copy a file named BC450rtl.dll manually from where you decompressed AN_WIN file, to ANFLOW.WIN directory. This thing was done intentionally.

    9) Now, you may run ANFLOW.

I have a beeping problem on my AN-2000.
I get rid of it only by turning the power OFF and ON.
The AN-2000 should not beep!!! It does only a small beep when it resets the CPU (the short time between power-on to software reset of all peripherals on the CPU board). The circuit in the new mainframe has a new reset circuit with a shorter relaxation time than the old one. Please wait 1-2 seconds between ON to OFF and vice versa.

If you still get a continuous beep, it means that either you have some line voltage interruptions or that the connectors of the power supply are not clean. If so, you should put the AN-2000 on a table, with the front panel facing down, remove the CPU board, Test channel and the power supply, clean their contacts, let them dry and return them back to their slots.

I don't understand how the CAL-2000 fits into the picture? Does the calibration file automatically get used by the ANFLOW program?
Do I use the calibration file after the data has been stored?

     After you run ANFLOW and call LINFLOW from the RUN menu, open a calibration file with a new name. You should check that the serial port parameters are set for both: the calibrator and the AN-2000. Then, you have to set the calibration parameters in "Configuration" menu and after that you select:
CALIBRATION => AUTO => SINGLE WIRE.

When you press OK the calibrator will start accelerating and do the calibration. After the calibration is done, you should save the calibration data into the same file. You may look into the calibration curve by pressing the GRAPH button.


Then, from the main menu, you select CALIBRATION => SEND and answer YES to all the messages that follow. The channels in the calibration file have to be operated.

Now, all the polynomial coefficients and other parameters are stored in the non-volatile memory of the AN-2000. You should exit LINFLOW and in the channel configuration, you must set these channels as "velocity" channels. From here on, at all of the operations done by the software the channels will be treated as "velocity" channels (i.e. in SCOPE you will see m/sec, data acquisition will be in m/sec etc.) Please make sure that in CHANNEL CONFIGURATION you activated the VELOCITY and/or TEMPERATURE scales.

When I'm in the "c:\linflow" directory and type "linflow", I get a message indicating the program requires a 286|DOS-Extender.
Please explain.


In the configuration set in your PC, LINFLOW is not a stand alone program.
It should be activated from ANFLOW only, where ANFLOW runs it in A special mode transferring several parameters between programs. The RUN286 directory should be in your path (it's the DOS Extender).

When in ANFLOW program and request scope mode, scope of turbulence intensity, or other VIEW screens, I get "Loss of
data synchronization error" . What can I do?

The problem is with the DMA cable. Check all of the connections between the AN-2000 and the SPD500 box and the PC. Check wether the cable was damaged.

Why should I prefer your system over the competitor's?

There are many reasons to prefer AA LAB's system:
Our system has an optical isolation between the PC ground and the Analog ground in the AN-2000.
That means: lower noise, no spikes (like those you get when you move the mouse in other systems), no Ground loops (= oscillations if grounds are connected together), no sensor burning (PC sends a lot of 1V ptp noise spikes on the +5V supply voltage of the PC).

Linearized data in real time = only our system can show you correlations between Velocity and Temperature or Voltage in real time with a refresh rate better than that of a digital oscilloscope!!! You can also see FFT of VELOCITY in real time (without harmonics) and you also make Data Acquisition in Engineering Units (Ideal for flow control and feedback systems).

Only our system has Calibrated Low Pass Filters with 512 cutoff frequencies.
That means that the delay is the same on all channels within +/-2% !!! So you get true simultaneous sampling of all channels (using our built-in A/Ds).
In the competitor's filters you get a phase lag difference of 10% between channels (filters are not calibrated) so that usually causes a jitter of +/- 1 sample (because you usually measure at a sampling rate of about 10 times the highest frequency of your data ).With twice more you cannot reconstruct the shape of the signal accurately.
The Low-Pass filter calibrator is built-in, so you may calibrate by yourself at any time.

We have 14 bit A/Ds built into the channels. That means:
a) signal path is only few Cm. (and not 5-10 meters as in the competitor's system).
b) We guarantee 1 LSB of noise (due to the optical isolation).
c) No ground loops (because in the competitor's system all the ground lines of the system are connected via many coaxial cables {=many grounds} to the same ground of the PC).

Our software does much better than the competitor's software: It does 1D, 2D and 3D calibrations, it has a control for almost any traverse system using RS-232 ASCII commands (or Pulse & Direction using a special software driver), works with Windows 95 / 3.1, LabView, DOS + source codes for interfacing, has many advanced analysis tools (like 3D profiles) etc.

Our resistance decade has higher accuracy and it's calibrated!!!

What is the format of the data acquisition binary file?

The format of the Data Acquisition binary file is as follows:


FIELD SIZE TYPE (PC's "C"type)
fOffset 2 [bytes] int.
fieldNumber 2 [bytes] int.
ASCII text variable size ASCII text
fOffset 2 [bytes] int.
fOffset 2 [bytes] (second copy ) int.
recordsNumber 4 [bytes] long
fieldNumber 2 [bytes] (second copy ) int.
recordLength 2 [bytes] int.
channelsArray 2 * fieldNumber [bytes] int.[fieldNumber]
channelsInfoArray 110 * fieldNumber [bytes] CH INFO[fieldNumber]
binaryData recordLength * recordsNumber [bytes] unsigned

 

 

 

 

 

 

 

 

 

 

FIELD OFFSET FIELD DESCRIPTION
Offset 0 offset of the 6 bytes + ASCII text size, this offset pointed to the second copy of the fOffset
fOffset 2 size of the header, after this offset binary data begins
fieldNumber 4 number of the sampled channels also used as a fields number in the RECORDS ASCII text
fOffset Offset size of the header, after this offset binary data begins (second copy)
recordsNumber Offset + 2 total number of the RECORDS in the file
fieldNumber Offset + 6 number of the sampled channels also used as a fields number in the RECORDS ( second copy )
recordLength Offset + 8 RECORDS length in bytes, fieldNumber * 2
channelsArray Offset + 10 array of the integers, each integer assumes the channel's number. The order of the channels (smallest number first) describes the order of the data in the RECORDS
channelsInfoArray Offset + 10 + array of the CH_INFO structures, array fieldNumber*2

contains CH_INFO structure for each channel presented in the channelsArray

binaryData fOffset this offset pointed to the first RECORDS entry


   All offsets are given as absolute offset from the beginning of the file.

   RECORDS is the continuous sampling data for all sampled channels; channels order in the RECORDS is described in the channelsArray field.

   Binary data is stored in "Two's Complement" A/D format for Voltage channels and linearized integer for the Temperature or Velocity channels. You can use the A2D_to_Double() "C" function to convert RECORDS data to the double format. Parameters:
- unsigned unsignedValue, data from the RECORDS.
- int. is_CH_LIN_READY, FALSE if current channel has Voltage type, TRUE otherwise.

   Unsigned TwosCompliment_2_PC (unsigned Val) {converts word from two's compliment to standard PC format
unsigned int compVal = 1 << 13;) bitwise shift left, moves the bits to the left, it discards the far left bit and assigns the right most bit to 0. After this operation compVal has 1 in bit 14, any other bits are 0. if( Val & compVal ) check, is number negative ?
return( Val & 0x1FFF );// number is negative because MSB is 1
else return( Val + 0x1FFF ); }

   double A2D_to_Double( unsigned unsignedValue, int. is_CH_LIN_READY )
{_ // converts from A/D number to double format
double doubleValue;
if( is_CH_LIN_READY )
doubleValue = A * unsignedValue + B;
else
doubleValue = A * TwosCompliment_2_PC( unsignedValue ) + B;
return doubleValue; }

   struct CH_INFO {
int. status;
// use A and B coefficients to convert from A/D number to desired unit.
double A;
double B;
int numberID;__ // channel number
int unitID;
double coefficients[ 7 ];__ // polynomial coefficients
// Velocity and Voltage boundary. Insize this range polynomial exists.
double VelocityMin;
double VelocityMax;
double VoltageMin;
double VoltageMax; };

   A and B parameters are in the "CH_INFO" structure in the header of the binary
and ASCII file. It is recommended to compare your results with a binary file
converted to ASCII by our "CONVERT" function in ANFLOW.

   Calculation example:
We read the binary data from the RECORD in the data file. The channel
has sampled this data, operated as a Voltage channel.
A=0.0012207,
B=-10.0.
The A/D voltage value is 0x3FEC, converting from two's complement yields:
If bit 14 of the A/D value = 1, number is negative, bitwise AND with 0x1FFF can
be
PC's value = A/D value & 0x1FFF => PC's value = 0x1FEC
otherwise, simply add 0x1FFF to A/D value:
PC's value = A/D value + 0x1FFF

For voltage calculation, A and B coefficients may be used:
Voltage = A * PC's value + B; voltage = -0.0244151 [V]

For velocity calculation, A = 0.00216071 and B = -4.9 coefficients may be used.
Velocity = A * A/D value + B; velocity = 30.45785844 [m/sec]
NOTE
A and B coefficients are different from one file to another,
they’re value-dependent to offset, gain, units and so on.

How much do I pay for an AN-2000 system?
In order to tell you "How much", we need to know some more details:

  What kind of measurement do you want to do with it?
             * Type of flow ?
             * Velocity range (approx.)?
             * Expected Turbulence level ?
             * Frequency response ?
             * Type of sensors you want to use ?
             * Any physical demands (distance from PC to experiment etc.) ?
             * Required S/N level ?

               Do you want to measure Temperature (ambient or fluctuations)?

               Do you want to monitor the Humidity?

               Are you driving a traverse while doing measurements?
                 _ (if yes - what interface? RS-232 or TTL Pulse & Direction?)
               Do you need external synchronization for ensemble averaging (for rotating machinery or for excited flow)? If yes - what is                  your input?

               What sampling rate do you need per channel (simultaneous)? 93 KHz or 160 KHz.?

               Do you need to sample any analog input (or an external transducer - pressure, strain, microphone etc.) together with your                  velocity signals?


                As you may see, because the instrument has over 19 different options, we must tailor it to your specific needs.


What upgrades can you offer for the ANFLOW software?
We offer the following upgrades:

             Windows 3.1 & 95 revisions.
             LABVIEW interfaces to the AN-2000.
             Traverse system control (for ANY traverse using RS-232 or Pulse & Direction TTL signals).
             3D and 2D calibration, plus real-time display of data.
             Automatic frequency response tuning (Optimization).
             Hardware module for multiplying the throughput (to memory) of our interface card.
             Probe2Audio software that enables you to record speech using your Hot-wire sensor... and play it in real time on a Sound                Blaster
card !(They cannot do that, because they do not process on-line linearization like we do)
             Printing of graphs to over 15 different printers and file formats.
             PROFLOW software for 2D and 3D velocity profiles.
             Temperature and Humidity recording for compensation of velocity.

     And much more.....


NOTE
We have a small add-on unit designated GAGE4000, which may be used for measuring ambient temperature (using a thermocouple sensor) and Humidity. This unit is attaches to the AN-2000 via the RS-232 cable and measures those parameters (or one of them) whenever you save data to file. This data (saved in ASCII into the header) is needed for Temp and RH compensation of your data. The price of this unit, including the two sensors, cables and manual is only US$480.- and it can also be operated from LABVIEW as a stand alone measurement and control system. It has three digital I/Os - one of them might be configured as a counter, PWM or Analog output, Excitation 0V +/- 5V, three fast A/D inputs (8 bits), two general purpose A/D inputs (18 bits, 1 Hz.), one Thermocouple input, etc.

 

 

 

 

 

 


I have difficulty with my resistance measurements - what do I do?



The calibration in one of the channels has probably been corrupted. You can re-calibrate that channel by performing the calibration procedure below:

1) Prepare two reference resistors soldered to a pair of MALE BNC plugs. The total resistances of these resistors (including the BNC contact) should be 0.5 and 5 ohms. Please use a high accuracy multi meter capable of measuring 0.01 ohms with +/- 0.005 ohms accuracy.

2) In the AN-2000 software select SETUP => SYSTEM => AUTO R COEF, then enter the number of channel you want to calibrate. Press CONTINUE to confirm.

3) The system will prompt you eight times to connect both resistance values to the four Bridge ratios(total of eight combinations). Connect the resistors you have prepared, according to the instructions on the screen, DIRECTLY to the PROBE BNC of the channel.

4)The system will ask you to switch the channel to one of the four bridge ratios: 1:1,1:2,1:10,1:20.
These bridge ratios are set using the two switches on the back of the channel module: POWER and BR. RATIO, where the POWER switch is used to switch between 1:1 or 1:10 to 1:2 or 1:20 using the NORMAL and HIGH positions, respectively.

5) After the calibration is done, exit the SETUP menu and reset the instrument.


I am experiencing some trouble with my RS-232 communication channel. What are the possible solutions?

Listed below are suggested solutions to the common problems with RS-232 communication.

The serial communication in the AN-2000 CPU and in the ANFLOW software are done via a 16550 UART (communication IC). There are several things which might cause an RS-232 communication error ("Error # xx, RS-232 transmission error" or "AN-2000 is not connected" etc.):

The serial port parameters are not set the same on both sides (i.e. on the ANFLOW software and in the AN-2000 internal setup).

Solution:Please follow the instructions in the manual.

Bad cable. The RS-232 cable is not connected, or the cable has a cut in the middle or the cable has been pulled out of either of the connectors.
Solution: Replace the cable by using a 1:1 cable wiring (pin#1 to pin#1, pin#2 to pin#2 etc.)

The RS-232 fault came after a power failure / short circuit, or computer memory/disk problem. That probably caused the corruption of the non-volatile memory in the AN-2000, the UART setup or the rs_setup.cfg file in the ANFLOW directory.

Solution:
   1) Turn off the power of the AN-2000. Remove the CPU module out of its slot for 1 minute, then return it. That will erase the non-volatile memory, and reset it to the default parameters.
   2) Erase the files C:\ANFLOW\rs_setup.cfg and C:\LINFLOW\rs_setup.cfg .
Turn off the PC (to reset the UART IC). Turn on the PC and the AN-2000 and run the software. The software and the AN-2000 will run on default parameters (9600,n,8,1,p through the port in the PC card, using IRQ# 5).
   3) If there is any conflict with another card in the PC, or the cable has been connected to COM1, COM2 or the card using IRQ#10, you will get the error message again. If you received that error message, you will need to set the RS-232 parameters again (both in ANFLOW and in LINFLOW).


Remove any resident programs in your AUTOEXEC.BAT and CONFIG.SYS files.
REMEMBER: you should reboot the PC after removing those programs. Please refer to the manual for more information.

Please check if the Motherboard of your PC or the BIOS setups were changed since the last time it worked. Sometimes, those parameters should influence the communication of the software with the PC card. You may also try to run the card with the software on another PC.

Did you install a modem / SCSI /sound blaster card recently? there might be an address/IRQ conflict. Try to remove those cards. If it solves the problem, change their configuration and try again

If all of the above doesn't help - please call us.

 

Common Operation Problems

When I run ANFLOW from DOS, the system "crashed" frequently or the computer hangs during Data Acquisition/Scope.
What is the problem?


These are problems often caused by collision of two interrupts in your computer or devices running on the same address. If you have a SCSI adapter BUS MOUSE another special card (like A/D GPIB cards) to check their addresses and interrupts.

The easiest thing will be to remove all the other cards and device drivers (from Autoexec and Config files) and then check again. If the problem disappeared - return one card and check again. When you have found the card which causes the malfunction - change its Address/IRQ to another available resource.

If the problem did not disappear - check the computer's BIOS setup to see which interrupts it uses
(Plug & Play or PCI ?).



I get an "Anti Aliasing error" message - what do I do?

Set the Low-pass filter in each channel to be less than 50% of the sampling rate. The message should disappear. This is an Anti Aliasing error caused by your sampling frequency being too low (Nyquist Theorem).



I have an Operate/Standby problem - what do I do?

This is an old bug. Please check if you are using an updated version of the Firmware.
Anyway, to work with it, you simply have to reset the instrument and "re operate" all channels while pressing MANUAL for cable and probe resistance. Since the AN-2000 recalls the old values, it will work the same way.



I have a scroll scope problem - the signal looks "broken" on the screen. What do I do?

You have an undersampling situation. Try to enter a sine wave of 0.5 Hz or less - and you'll see the difference!


I do not see the FFT graph in FFT screen - what do I do?

If you see a good signal on "SCOPE" screen, you should be able to see the FFT on the screen. Try to play with the scales. If you sample at a high sampling rate with no flow, the graph looks "empty" because of the very low noise of the A/D's.



Can I change/reverse the scales in LINFLOW? Why does the graph show Voltage versus Speed?

It's not possible to change the scale, unless you pick the "Zoom" tool and mark an area on the screen (using the mouse). The graph shows the points of the calibration as they are collected, while the input to your system is the Velocity ("X") and the output is digitized Voltage ("Y").




Common CCA* Questions
(*CCA=Constant Current Anemometry)
I see a lot of 60, 120, 180 Hz noise on my CCA signal. What's wrong? (50, 100, 150 Hz. for Europe and Asia).

The 60 Hz signal that you see (in any anemometer) in CCA mode (only!!!) and its harmonics (120Hz, 180Hz. etc. ) are caused by the physical nature of the CCA circuit. The CCA does not have a feedback loop (like CTA) which cancels all such noises. It is actually very similar to an antenna amplifier (a current source that feeds the "Antenna" and a voltage amplifier).
Therefore, it collects all the noises from the environment (including 60 Hz. of course)and high frequency noises as well.

Solution:
   1) Take the cable away from any transformers, motors etc.
   2) Use a new cable with good shielding (also try magnetic shielding).
   3) Remove and transfer the internal power supply to a separate mainframe, by using the Opt. 13 power supply ext. mounting kit ($435).
   4) Try to ground one side of the cable to an "Earth" ground.



There is a "drift" of 0.3 V at the CCA channels during 24 Hours. Is this O.K.?

Consider the sensitivity of your CCA channel - it depends on the sensor current, sensor's "Alfa", Gain etc. Check what TEMPERATURE SCALE you are using - it could be that 0.3V is equal to 0.01deg.C, in some cases. Check the drift and accuracy parameters of your thermocouple and reference instrument.



I am trying to work at an OHR of 1.01 and measure temperature. Why do I get sensitivity to velocity?

About working with OHR of 1.01: in theory it is possible, but it is not accurate. Use it only if you have no other choice. The velocity sensitivity that you received is because you had some error in the adjustment of the sensor or cable resistance. Try to do it again or try OHR of 1.00 or even 0.99 (to compensate for the adjustment errors).




For more information please contact us at:

Head Office: 33 Hayetzira St., Ramat-Gan 52521, ISRAEL. Tel: 972-3-5756327, Fax: 972-3-5756326
U.S.A office: 3607 Silverside Rd., Wilmington, DE19810. Tel: (302) 478-2881, Fax: (302) 478-7956
E-mail: aalabs@lab-systems.com



Copyright 2000, A.A.Lab Systems Ltd.

Back to Home Page    Products Page    Sales Page    Services Page    Technical Support Page    Company Profile Page    No JavaScript Page